Benchmarking Action Spaces in Reinforcement Learning for Vision-based Robotic Manipulation

• URL: http://arxiv.org/abs/2606.18594v1
• Date: Wed, 17 Jun 2026 01:45:13 GMT
• Title: Benchmarking Action Spaces in Reinforcement Learning for Vision-based Robotic Manipulation
• Abstract summary: In real-world reinforcement learning (RL), the choice of action space can play a key role in shaping motion smoothness, safety, and overall task performance.<n>We train policies in simulation and deploy them to the real world using sim-to-real transfer.<n>We find that action-space representation indeed significantly affects sim-to-real performance.
• Score: 9.501565892264027
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: In real-world reinforcement learning (RL), the choice of action space can play a key role in shaping motion smoothness, safety, and overall task performance. In this study, we evaluate pose increment, pose velocity, joint position increment, and joint velocity across two vision-based manipulation tasks: object picking and pushing. We train policies in simulation and deploy them to the real world using sim-to-real transfer. We find that action-space representation indeed significantly affects sim-to-real performance. In particular, we find that the joint velocity action space is best for the vision-based picking and pushing tasks in terms of smoothness and final task performance. We also provide practical guidance for RL practitioners in choosing action spaces for both simulation and real-world experiments.

Related papers

• Exploratory, Communicative, and Deployable: Vision-Driven Embodied Agents for Open-World Mobile Manipulation [47.58516950734307]
  Existing frameworks often rely on privileged simulator states or assume complete instructions, bypassing realistic deployment challenges.<n>We present REAL, an agentic framework for open-world mobile manipulation.<n>We design diverse task compositions to drive data collection, supervised fine-tuning, and online reinforcement learning.
  arXiv  Detail & Related papers  (2026-07-15T09:55:45Z)
• Universal Pose Pretraining for Generalizable Vision-Language-Action Policies [83.39008378156647]
  Existing Vision-Language-Action (VLA) models often suffer from feature collapse and low training efficiency.<n>We propose Pose-VLA, a decoupled paradigm that separates VLA training into a pre-training phase for extracting universal 3D spatial priors.<n>Our framework follows a two-stage pre-training pipeline, establishing fundamental spatial grounding via poses followed by motion alignment.
  arXiv  Detail & Related papers  (2026-02-23T11:00:08Z)
• PhysHSI: Towards a Real-World Generalizable and Natural Humanoid-Scene Interaction System [67.2851799763138]
  PhysHSI comprises a simulation training pipeline and a real-world deployment system.<n>In simulation, we adopt adversarial motion prior-based policy learning to imitate natural humanoid-scene interaction data.<n>For real-world deployment, we introduce a coarse-to-fine object localization module that combines LiDAR and camera inputs.
  arXiv  Detail & Related papers  (2025-10-13T07:11:37Z)
• Watch Less, Feel More: Sim-to-Real RL for Generalizable Articulated Object Manipulation via Motion Adaptation and Impedance Control [7.986465090160508]
  We present a novel RL-based pipeline equipped with variable impedance control and motion adaptation.<n>Our pipeline focuses on smooth and dexterous motion during zero-shot sim-to-real transfer.<n>To the best of our knowledge, our policy is the first to report 84% success rate in the real world.
  arXiv  Detail & Related papers  (2025-02-20T11:18:35Z)
• On the Role of the Action Space in Robot Manipulation Learning and Sim-to-Real Transfer [6.1622717998840395]
  We study the choice of action space in robot manipulation learning and sim-to-real transfer. We train over 250 reinforcement learning(RL) agents in simulated reaching and pushing tasks. The choice of spaces spans combinations of common action space design characteristics.
  arXiv  Detail & Related papers  (2023-12-06T18:38:05Z)
• Universal Humanoid Motion Representations for Physics-Based Control [71.46142106079292]
  We present a universal motion representation that encompasses a comprehensive range of motor skills for physics-based humanoid control. We first learn a motion imitator that can imitate all of human motion from a large, unstructured motion dataset. We then create our motion representation by distilling skills directly from the imitator.
  arXiv  Detail & Related papers  (2023-10-06T20:48:43Z)
• Contrastive Learning for Enhancing Robust Scene Transfer in Vision-based Agile Flight [21.728935597793473]
  This work proposes an adaptive multi-pair contrastive learning strategy for visual representation learning that enables zero-shot scene transfer and real-world deployment. We demonstrate the performance of our approach on the task of agile, vision-based quadrotor flight.
  arXiv  Detail & Related papers  (2023-09-18T15:25:59Z)
• Sim-to-Real Deep Reinforcement Learning with Manipulators for Pick-and-place [1.7478203318226313]
  When transferring a Deep Reinforcement Learning model from simulation to the real world, the performance could be unsatisfactory. This paper proposes a self-supervised vision-based DRL method that allows robots to pick and place objects effectively.
  arXiv  Detail & Related papers  (2023-09-17T11:51:18Z)
• Nonprehensile Riemannian Motion Predictive Control [57.295751294224765]
  We introduce a novel Real-to-Sim reward analysis technique to reliably imagine and predict the outcome of taking possible actions for a real robotic platform. We produce a closed-loop controller to reactively push objects in a continuous action space. We observe that RMPC is robust in cluttered as well as occluded environments and outperforms the baselines.
  arXiv  Detail & Related papers  (2021-11-15T18:50:04Z)
• ReLMoGen: Leveraging Motion Generation in Reinforcement Learning for Mobile Manipulation [99.2543521972137]
  ReLMoGen is a framework that combines a learned policy to predict subgoals and a motion generator to plan and execute the motion needed to reach these subgoals. Our method is benchmarked on a diverse set of seven robotics tasks in photo-realistic simulation environments. ReLMoGen shows outstanding transferability between different motion generators at test time, indicating a great potential to transfer to real robots.
  arXiv  Detail & Related papers  (2020-08-18T08:05:15Z)
• Point Cloud Based Reinforcement Learning for Sim-to-Real and Partial Observability in Visual Navigation [62.22058066456076]
  Reinforcement Learning (RL) represents powerful tools to solve complex robotic tasks. RL does not work directly in the real-world, which is known as the sim-to-real transfer problem. We propose a method that learns on an observation space constructed by point clouds and environment randomization.
  arXiv  Detail & Related papers  (2020-07-27T17:46:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.